DNA is a linear information storage medium, like a magnetic tape, but written with a very limited character set—just A, C, G, and T.
Here’s how we read that tape today: we dub a copy of the tape, stop at some point when we reach the character A, cut the tape there, and toss the fragment into the A bin. Then we make another copy, stopping at some point on another A, snip the new copy short, and toss that fragment into the bin. We repeat this millions of times, and then start over again for C. And again for G and T. When we’re done, we measure lengths of all of the millions of A fragments, C fragments, G fragments, and T fragments and then use a computer to sort them by length and tell us where in the sequence we encountered all of the As, Cs, Gs, and Ts. What could go wrong?
It would obviously be much quicker, cheaper, and more reliable if we had a DNA-read-head, a device we could just run the DNA-tape over to detect the sequence directly, in a single pass. Back in 1995, George Church and colleagues figured out one approach: it might be possible to use a low intensity electric current to pull long strands of DNA through nanometer-scale pores in a membrane and measure the electric field variations of the four nucleic acids—A, C, G, T—as they passed through.
Scientists are working on it, but we’re not there yet. Key questions remain unanswered. One of them is fundamental: How does DNA move through a pore? Does it slide through end-on? Or does the pore grab it somewhere in between, bend it double, and suck it through doubled over (like a strand of spaghetti slurped up from the middle). Every cell in your body, after all, carries about two or three meters of DNA. Even a mere fragment of 50,000 base pairs is about 16.5 micrometers long—thousands of times the diameter of a nanopore. If mere chance determines the orientation, one would expect that almost all DNA should pass through a nanopore doubled over.
Fortunately for the future of bedside genomics, that doesn’t seem to be the case. Researchers at Brown University have developed an elegant method for determining exactly how a DNA molecules passes through a nanopore. They can see if it slips through end-on or goes through doubled over…and, if it does double over, they can tell where along its length it folds.
Physicists Mirna Mihovilovic, Nick Hagerty, and Derek Stein followed the travels of about 1100 pieces of double-stranded DNA (for aficionados, they were 48,502 base-pair segments of bacteriophage lambda, measuring about 16.5 micrometers in length) as they slithered through an 8-nanometer-diameter pore in a 20-nm-thick silicon nitride membrane, urged along by a 3.6 nanoampere current. They measured the current flowing through the pore 50,000 times a second, and found it dropped when passing DNA partially blocked the pore. The interruptions lasted only a couple of milliseconds, and the magnitude of the current reduction was proportional to the cross-section of DNA blocking the pore—which is to say that the current dropped about 0.28 nanoamperes when there was a single strand in the pore, and about 0.56 nA when the DNA passed through “sideways.” Thus, the current profile revealed the DNA’s orientation: the relative durations of the double-strand and single-strand current reductions showed just where the molecule had folded. (So a 3-millisecond drop of 0.28 nA might indicate that the DNA had speared straight through the pore, while a 1.5-millisecond drop of 0.56 nA indicated a fold exactly in the middle. The diagram at right makes it clearer: ECD stands for “event charge deficit,” the current drop integrated over time, which remains approximately constant for each DNA passage.)
The researchers found that DNA passes through the pore smoothly, end-on, about 25% of the time—far more than most current models would predict. This surprisingly high proportion—indicating, they say that the orientation is a function of “the configurational entropy of the approaching polymer”—bodes well for developing a nanopore-based DNA direct reader.
Figures: Derek Stein/Brown University